The challenges of assessing adiposity in a clinical setting
临床环境中评估肥胖的挑战

Obesity increases the risks of developing cardiometabolic disease, and reduces life expectancy by approximately 5 to 20 years¹. However, increased adiposity might not be the only driver of this increased risk of disease, as body adipose tissue distribution, specifically elevated visceral adipose tissue (VAT) mass, influences disease risk independently of total weight². Although initially regarded as a homogeneous tissue, adipose is now recognized as an important endocrine organ with depot-specific differences (Box 1). It is particularly noteworthy that subcutaneous and visceral white adipose tissues respond differently to metabolic challenges, with variations in endothelial function, lipid turnover and susceptibility to vasculature inflammation^3,4,5. Furthermore, increased VAT mass, relative to subcutaneous adipose tissue (SAT) mass, is associated with an elevated risk of developing cardiometabolic conditions, such as type 2 diabetes mellitus (T2DM), hypertension and cardiovascular disease⁶. Understanding the individual patient’s adipose tissue distribution can thus provide critical insight into obesity-related pathophysiology and aid therapeutic approaches, but this endeavour requires accurate clinical measurements of body adipose tissue and its distribution.
肥胖增加了患心血管代谢疾病的风险，并使预期寿命减少约 5 至 20 年 ¹ 。然而，体脂增加可能并非疾病风险增加的唯一驱动因素，因为身体脂肪组织分布，特别是内脏脂肪组织（VAT）质量的增加，独立于总体重影响疾病风险 ² 。尽管最初被视为同质组织，脂肪现在被认识到是一个重要的内分泌器官，具有部位特异性差异（见框 1）。特别值得注意的是，皮下和内脏白色脂肪组织对代谢挑战的反应不同，表现在内皮功能、脂质周转和对血管炎症的易感性上的差异 ^3,4,5 。此外，相对于皮下脂肪组织（SAT）质量，VAT 质量的增加与患心血管代谢疾病（如 2 型糖尿病（T2DM）、高血压和心血管疾病）的风险升高相关 ⁶ 。 了解个体患者的脂肪组织分布因此能为肥胖相关病理生理学提供关键见解，并辅助治疗方法，但这一努力需要准确临床测量身体脂肪组织及其分布。

However, performing accurate adiposity measurements is challenging, particularly in a clinical research setting. There is no one-size-fits-all method that accurately captures an individual’s total body adiposity and adipose tissue distribution. Various measurement techniques, such as BMI, bioelectrical impedance analysis (BIA), dual-energy X-ray absorptiometry (DXA) and MRI, all provide different insights⁷ (Tables 1,2), making it difficult to choose a single, universally applicable method. Many measurements are influenced by hydration, muscle mass or bone density, leading to potential inaccuracies, while cost and accessibility can limit the use of advanced imaging techniques in resource-constrained clinical settings. The accuracy of each measurement method is also a critical concern⁸, as treatment plans and weight loss tracking rely heavily on precision. This Perspective article discusses the importance of carefully selecting and interpreting adiposity and adipose tissue distribution measurements, the strengths and weaknesses of the respective approaches, and how these challenges can influence our understanding and interpretation of research data.
然而，进行准确的肥胖测量具有挑战性，特别是在临床研究环境中。没有一种放之四海而皆准的方法能够精确捕捉个体的总体肥胖程度和脂肪组织分布。多种测量技术，如 BMI、生物电阻抗分析（BIA）、双能 X 射线吸收法（DXA）和 MRI，各自提供了不同的见解 ⁷ （表 1、2），这使得选择一种普遍适用的单一方法变得困难。许多测量结果受水分、肌肉量或骨密度的影响，可能导致不准确，而成本和可及性则限制了在资源有限的临床环境中使用先进的成像技术。每种测量方法的准确性也是一个关键问题 ⁸ ，因为治疗计划和减重追踪高度依赖于精确性。本文讨论了精心选择和解释肥胖及脂肪组织分布测量的重要性、各自方法的优缺点，以及这些挑战如何影响我们对研究数据的理解和解释。

The function of adipose tissue differs depending on its anatomical location (Box 1; Fig. 1). In addition, adipose tissue function and distribution can vary between children and adults, people of different ethnic backgrounds, and women and men. This variation calls for a need to apply age-specific, ethnicity-specific and sex-specific standards when analysing body composition, as these parameters can influence clinical interpretation and the risk of developing obesity-related diseases and comorbidities.
脂肪组织的功能因其解剖位置而异（框 1；图 1）。此外，脂肪组织的功能和分布在儿童与成人、不同种族背景的人群以及女性和男性之间可能存在差异。这种差异要求在分析身体成分时应用年龄特定、种族特定和性别特定的标准，因为这些参数可能影响临床解释以及发展肥胖相关疾病和合并症的风险。

a, Cellular morphology and function of different adipocyte types. Human adipocytes can be classified as white, brown or beige. White adipocytes are the most prevalent and contain a large unilocular triglyceride droplet, enabling long-term energy storage and insulation. Brown adipocytes, primarily found in the neck and upper back, are densely packed with mitochondria, giving the cell a brownish colour and a uniquely high capacity for thermogenesis. Beige adipocytes are rare, but their presence can increase due to certain stimuli, such as cold exposure or exercise. The beige adipocytes share characteristics of both white and brown adipocytes and have thermogenic capacity that can contribute to energy expenditure. The white adipose tissue depots are the largest and most relevant for the clinical quantification of adipose tissue distribution. b, Anatomical depiction of white adipose tissue and ectopic fat depots. White adipose tissue is divided into the subcutaneous adipose tissue (SAT) and visceral adipose tissues (VAT). In humans, SAT tends to accumulate around the abdomen, thighs and buttocks. The abdominal adipose tissue can be divided into a superficial and a deep subcutaneous layer. The VAT is divided into the omentum, which hangs as an ‘apron’ from the stomach to the liver, protecting the visceral organs, and intestinal mesenteric adipose tissue, which upholds the intestinal barrier⁶. If the adipose compartments surpass their physiological lipid storage capacity, triglycerides accumulate in other organs as ectopic adipose tissue, which refers to lipid deposition in the liver, epicardium, pancreas and skeletal muscles.
a, 不同脂肪细胞类型的细胞形态与功能。人类脂肪细胞可分为白色、棕色或米色。白色脂肪细胞最为普遍，内含一个大的单房性甘油三酯液滴，能够长期储存能量并提供保温作用。棕色脂肪细胞主要分布于颈部和上背部，富含线粒体，赋予细胞棕色外观及独特的高产热能力。米色脂肪细胞较为罕见，但在特定刺激下（如寒冷暴露或运动）其数量可增加。米色脂肪细胞兼具白色和棕色脂肪细胞的特征，具备产热能力，有助于能量消耗。白色脂肪组织库是最大且对临床量化脂肪组织分布最为重要的部分。b, 白色脂肪组织及异位脂肪库的解剖示意图。白色脂肪组织分为皮下脂肪组织（SAT）和内脏脂肪组织（VAT）。在人体中，SAT 倾向于积聚在腹部、大腿和臀部周围。 腹部脂肪组织可分为浅层和深层皮下脂肪层。内脏脂肪组织（VAT）分为网膜，它像“围裙”一样从胃悬挂至肝脏，保护内脏器官，以及肠系膜脂肪组织，它支撑着肠道屏障 ⁶ 。如果脂肪隔室超过其生理脂质储存能力，甘油三酯会在其他器官中积累，形成异位脂肪组织，这指的是肝脏、心外膜、胰腺和骨骼肌中的脂质沉积。

Full size image

Adiposity differences in children versus adults
儿童与成人之间的肥胖差异

The prevalence of childhood obesity remains substantial^9,10,11, and as mothers can genetically ‘imprint’ the risk of developing obesity onto their children, the disease burden can span generations^12,13.
儿童肥胖的患病率仍然居高不下 ^9,10,11 ，由于母亲能够通过基因将肥胖风险“印记”给子女，这种疾病负担可能跨越数代 ^12,13 。

Historically, infant size and shape are quantified by standard anthropometry, beginning with birthweight¹⁴. However, relatively little is understood about body composition during infancy. Importantly, the first 1,000 days (that is, from conception to the age of 2 years) represent a ‘critical window’ that can affect growth and development, and potentially also body composition¹⁵. A global consensus on definitions and protocols for body composition assessment in infancy would thus be most valuable to the field¹⁶. Indeed, recent attempts to assemble growth and body composition trajectories in infants have been made (for example, in a 2022 study using air displacement plethysmography¹⁷). Nonetheless, more research is needed to determine how infant weight and adipose tissue distribution affects long-term trends and predictive outcomes of an individual’s risk of developing cardiometabolic disease.
历史上，婴儿的体型和形态通过标准人体测量学进行量化，始于出生体重 ¹⁴ 。然而，对于婴儿期的身体组成了解相对较少。重要的是，最初的 1000 天（即从受孕到 2 岁）代表了一个“关键窗口”，可以影响生长和发育，并可能影响身体组成 ¹⁵ 。因此，关于婴儿身体组成评估的定义和协议的全球共识对该领域将极为宝贵 ¹⁶ 。事实上，最近已经尝试汇编婴儿的生长和身体组成轨迹（例如，在 2022 年的一项使用空气置换体积描记法的研究 ¹⁷ ）。尽管如此，还需要更多的研究来确定婴儿体重和脂肪组织分布如何影响长期趋势以及个体发展心脏代谢疾病风险的预测结果。

In children, weight gain is usually based on fat-free mass, rather than fat mass, as the proportion of adipose tissue mass declines during childhood, at least until puberty. It is challenging to assess body composition alterations related to children’s growth as it occurs in spurts¹⁸, but in adolescence, a higher BMI is associated with increased risk of disease development later in life (for example, T2DM and cardiometabolic disease)^19,20,21. Similar to the findings in adults, ethnicity and sex affect both obesity prevalence²² and adiposity distribution²³ in children.
在儿童中，体重增加通常基于无脂肪质量，而非脂肪质量，因为儿童期脂肪组织质量的比例下降，至少直到青春期。评估与儿童生长相关的身体成分变化具有挑战性，因为这种变化是间歇性的 ¹⁸ ，但在青春期，较高的 BMI 与日后疾病（如 2 型糖尿病和心脏代谢疾病）发展风险增加相关 ^19,20,21 。与成人的研究结果相似，种族和性别影响儿童的肥胖率 ²² 和脂肪分布 ²³ 。

Unfortunately, obesity research is poorly validated in children, and it remains unclear how the scientific community should handle this issue. For instance, what clinical guidelines are appropriate for weight-loss interventions in children? It is known that gastric bypass causes nutritional deficits in ;vitamin B₁₂, folic acid, iron, calcium and thiamine due to rearrangement of the gastrointestinal tract²⁴. Although bariatric surgery appears to be low-risk in adolescents^25,26, how a lifelong nutritional deficit and supplementation therapy affects an individual’s physiology and cardiometabolic status remains poorly understood, and more research in this field would be welcome.
遗憾的是，儿童肥胖研究的验证不足，科学界应如何处理这一问题仍不明确。例如，针对儿童减重干预的临床指南应如何制定？已知胃旁路手术因胃肠道重组会导致维生素 B ₁₂ 、叶酸、铁、钙和硫胺素的营养缺乏 ²⁴ 。尽管减重手术在青少年中似乎风险较低 ^25,26 ，但终身营养缺乏及补充疗法如何影响个体的生理和心脏代谢状态仍知之甚少，该领域亟需更多研究。

Ethnicity-related differences in adiposity and adipose tissue distribution
与种族相关的肥胖和脂肪组织分布差异

The prevalence of obesity-induced T2DM varies across different ethnic groups²⁷, with some of the most high-risk groups being Hispanic people, African people and some Asian subpopulations²⁸. Furthermore, T2DM onset occurs at a younger age and at lower BMI thresholds in African and Asian populations compared with European populations^29,30.
肥胖诱发的 2 型糖尿病（T2DM）的患病率在不同种族群体中存在差异 ²⁷ ，其中一些高风险群体包括西班牙裔人群、非洲裔人群以及某些亚洲亚群体 ²⁸ 。此外，与欧洲人群相比，非洲和亚洲人群的 T2DM 发病年龄更早，且 BMI 阈值更低 ^29,30 。

Body adipose tissue distribution patterns vary between ethnicities, and this variation might partly explain ethnic differences in obesity-associated disease risks. The underlying mechanisms are still not understood, but probably involve genetic traits that influence body adipose tissue distribution³⁰. Within the same obesity category, Asian individuals typically present with more VAT than African or European individuals, which might explain the higher risk of developing T2DM in some Asian populations compared with other ethnic groups^28,31,32. This could indicate that the SAT expansion capacity in individuals of Asian ethnicity is insufficient to support a high-calorie intake. Genome-wide association studies have identified loci that might genetically predispose ethnic differences in adipose tissue distribution and contribute to understanding ethnicity-dependent adipose tissue distribution diversity³¹. Interestingly, it is speculated that the ethnicity-specific differences in adipose tissue distribution and function could also be related to adipose tissue-induced systemic oxidative stress, and thus oxidative stress could contribute to the elevated susceptibility to metabolic diseases specifically in African women³³.
身体脂肪组织分布模式在不同种族之间存在差异，这种差异可能部分解释了与肥胖相关疾病风险的种族差异。其潜在机制尚不清楚，但可能涉及影响身体脂肪组织分布的遗传特征 ³⁰ 。在同一肥胖类别中，亚洲个体通常比非洲或欧洲个体拥有更多的内脏脂肪组织（VAT），这可能解释了某些亚洲人群相比其他种族群体发展 2 型糖尿病（T2DM）风险更高的原因 ^28,31,32 。这可能表明亚洲种族个体的皮下脂肪组织（SAT）扩展能力不足以支持高热量摄入。全基因组关联研究已经识别出可能遗传性地导致脂肪组织分布种族差异的基因位点，并有助于理解依赖于种族的脂肪组织分布多样性 ³¹ 。 有趣的是，据推测，脂肪组织分布和功能的种族特异性差异也可能与脂肪组织诱导的全身氧化应激有关，因此氧化应激可能特别增加了非洲女性对代谢疾病的易感性 ³³ 。

Importantly, randomized trials have shown that all ethnic groups benefit from weight loss; for example, the UK-based DiRECT trial including predominantly Europeans³⁴, the Qatar-based DIADEM-I trial including participants of Middle Eastern descent³⁵, and the STANDby trial including South Asian people with obesity³⁶. However, despite progress in the past decade, there is the prevailing problem of unreported race and/or ethnicity data of participants enrolled in studies, and the over-representation of white people in many risk-association studies and clinical trials³⁷.
重要的是，随机试验表明所有种族群体都能从减重中获益；例如，以欧洲人为主的英国 DiRECT 试验 ³⁴ ，包含中东裔参与者的卡塔尔 DIADEM-I 试验 ³⁵ ，以及针对南亚肥胖人群的 STANDby 试验 ³⁶ 。然而，尽管过去十年有所进展，研究中参与者种族和/或民族数据未报告的问题普遍存在，且许多风险关联研究和临床试验中白人比例过高 ³⁷ 。

To conclude, we highlight the fact that some ethnic groups are under-represented in body-composition investigations. Due to the relative lack of comparable data, it remains uncertain what methodology is the most appropriate for comparing adiposity across different ethnic groups³⁸.
综上所述，我们强调指出，某些族群在身体成分研究中的代表性不足。由于相对缺乏可比数据，目前尚不确定哪种方法最适合用于比较不同族群间的肥胖程度 ³⁸ 。

Sex differences in adipose tissue distribution and function
脂肪组织分布和功能的性别差异

Clinical observation and epidemiological evidence shows that women, at a population level, can live longer with overweight and obesity without developing T2DM than men^39,40. This finding is largely attributed to differences in adipose tissue distribution, as women have more subcutaneous and gluteal adipose tissue and a ‘pear-shaped’ adipose tissue distribution, while men have more VAT and an ‘apple-shaped’ adipose tissue distribution, which is associated with cardiometabolic risk factors^41,42. The underlying mechanisms behind altered adipose distribution in men and women are still not entirely understood but are probably multifactorial. Important contributors are sex hormones, cell-intrinsic factors, adipose depot microenvironments and tissue-specific genetic and epigenetic influences⁴³.
临床观察和流行病学证据表明，在群体水平上，女性在超重和肥胖状态下比男性更长寿且不易发展为 2 型糖尿病（T2DM） ^39,40 。这一发现主要归因于脂肪组织分布的差异，女性拥有更多的皮下和臀部脂肪组织，呈现“梨形”脂肪分布，而男性则更多内脏脂肪组织（VAT），呈现“苹果形”脂肪分布，后者与心脏代谢风险因素相关 ^41,42 。男女脂肪分布差异背后的机制尚未完全明了，但可能是多因素共同作用的结果，包括性激素、细胞内在因素、脂肪库微环境以及组织特异性遗传和表观遗传影响 ⁴³ 。

Sex differences in adiposity emerge during puberty and diminish at the onset of menopause. Indeed, menopause shifts the female adipose tissue function (increased adipocyte hypertrophy, immune cell infiltration and fibrosis) and adipose tissue distribution towards a more central obesity pattern resembling that of men⁴⁴, thereby demonstrating the importance that sex hormones play in adipose tissue distribution. Thus, biological actions of oestrogen and its receptor are a key component to sex differences in adipose distribution, and women with high testosterone levels (due to polycystic ovary syndrome) are at an increased risk of developing T2DM, compared with women with normal levels of testosterone, irrespective of age and BMI⁴⁵. However, androgen hormones are not necessarily responsible for poor metabolic outcomes, as testosterone deficiency due to hypogonadism increases men’s risk of developing T2DM and cardiovascular disease⁴⁶. Importantly, sex also has an effect on the overall ‘health’ of adipose tissues, which is not necessarily related to the depot-specific distribution differences between men and women. For instance, male adipose tissues are more likely to be inflamed and have increased fibrosis relative to female adipose tissues, regardless of distribution⁴², which, importantly, increases susceptibility to T2DM⁴⁷.
青春期期间，性别差异在脂肪分布上显现，并在更年期开始时减弱。实际上，更年期使女性脂肪组织功能（脂肪细胞肥大、免疫细胞浸润和纤维化增加）及脂肪分布向更类似于男性的中心性肥胖模式转变 ⁴⁴ ，这证明了性激素在脂肪分布中的重要性。因此，雌激素及其受体的生物学作用是脂肪分布性别差异的关键因素，且与睾酮水平正常的女性相比，睾酮水平高的女性（由于多囊卵巢综合征）不论年龄和 BMI 如何，患 2 型糖尿病（T2DM）的风险增加 ⁴⁵ 。然而，雄激素并不必然导致不良代谢结果，因为性腺功能减退导致的睾酮缺乏会增加男性患 T2DM 和心血管疾病的风险 ⁴⁶ 。重要的是，性别也影响脂肪组织的整体“健康”，这并不一定与男女之间特定脂肪库分布的差异相关。 例如，相对于女性脂肪组织，男性脂肪组织更容易发生炎症并增加纤维化，无论其分布如何 ⁴² ，这一点显著增加了对 2 型糖尿病（T2DM）的易感性 ⁴⁷ 。

Understanding the mechanisms that govern adipose distribution and sex differences in humans presents opportunities to advance the field of precision medicine. Such advances would be particularly useful for the early identification of at-risk individuals and the development of novel therapeutic strategies that alleviate central obesity and related cardiometabolic disorders⁴³. However, our understanding of mechanisms and underlying sex-dependent differences in adiposity and cardiometabolic disease risk is still developing. A prominent example is the finding that, although women with obesity live longer without developing hyperglycaemia than men, they are at a higher risk of exhibiting subclinical atherosclerosis if they progress to T2DM^39,40. Indeed, women with obesity who also have T2DM present more rapidly with cardiovascular risk factors than men, regardless of age or race^48,49,50,51. The greater impact of T2DM in women cannot be understated, particularly as the general awareness about the extent of cardiovascular disease in women remains underappreciated. We argue that sex differences should be a factor to consider when studying an individual’s risk of obesity and metabolic dysfunction. Sex is important for determining treatment strategies and clinical guidelines, as well as for providing opportunities for precision medicine.
理解调控人体脂肪分布及性别差异的机制，为精准医学领域的发展提供了契机。这些进展尤其有助于早期识别高危个体，并开发缓解中心性肥胖及相关心脏代谢疾病的新型治疗策略 ⁴³ 。然而，我们对于脂肪堆积及心脏代谢疾病风险中性别依赖性差异的机制理解仍在发展中。一个显著的例子是，尽管肥胖女性在未发展为高血糖的情况下比男性寿命更长，但若进展至 2 型糖尿病（T2DM），她们表现出亚临床动脉粥样硬化的风险更高 ^39,40 。事实上，患有 T2DM 的肥胖女性，无论年龄或种族，比男性更快出现心血管风险因素 ^48,49,50,51 。T2DM 对女性的更大影响不容忽视，尤其是考虑到公众对女性心血管疾病程度的认识仍显不足。 我们认为，在研究个体肥胖和代谢功能障碍风险时，性别差异应作为一个考虑因素。性别对于确定治疗策略和临床指南至关重要，同时也为精准医学提供了机会。

A range of tools and methods have been developed to quantify adiposity and body composition (Fig. 2; Tables 1,2). Each method has advantages and limitations, and their appropriateness depends on factors such as the level of accuracy required, accessibility, and the specific population being studied. Although practical to implement, BMI and other anthropometric surrogate measures are relatively poor predictors of an individual’s adipose tissue distribution and metabolic risk; therefore, body composition measurements that assess the relative proportion and placement of adipose tissue, muscle and bone are often considered more valuable. The choice usually comes down to a trade-off between cost and availability. To choose the most adequate method, and importantly, to interpret data correctly, it is essential to discuss the strengths and weaknesses of the respective methods (Tables 1,2). This can be particularly relevant for large clinical trials and weight-loss interventions (Supplementary Box 1). Indeed, large clinical trials are often unable to implement the use of accurate but expensive diagnostic imaging methods, such as DXA or MRI. In addition, substantial weight loss can alter body composition to the extent that certain measurements become less accurate (for example, due to belly and skin displacement), as outlined in this section.
一系列工具和方法已被开发用于量化肥胖和身体成分（图 2；表 1、2）。每种方法都有其优势和局限性，其适用性取决于所需精度水平、可及性以及研究的具体人群等因素。尽管 BMI 和其他人体测量替代指标实施起来较为方便，但在预测个体脂肪组织分布和代谢风险方面相对较差；因此，评估脂肪组织、肌肉和骨骼相对比例及分布的身体成分测量通常被认为更具价值。选择通常归结为成本与可用性之间的权衡。为了选择最合适的方法，并正确解读数据，讨论各自方法的优缺点至关重要（表 1、2）。这对于大型临床试验和减重干预尤其相关（补充框 1）。实际上，大型临床试验往往无法采用准确但昂贵的诊断成像方法，如 DXA 或 MRI。 此外，如本节所述，显著的体重减轻可能会改变身体成分，以至于某些测量结果变得不太准确（例如，由于腹部和皮肤位移）。

Summary of anthropometric measures and indices, as well as instrumentation-based tools to assess adipose tissue content and distribution. The left-hand panel summarizes the most commonly used anthropometric measures, such as BMI, waist-based and neck-based circumference measurements and sagittal abdominal diameter (SAD). It also provides the calculations utilized for adiposity indices. The right-hand panel outlines instrumentation-based tools used to assess adipose tissue content and distribution. These tools include BIA, as well as air-displacement and water-displacement methods, and imaging techniques such as DXA, CT and MRI. The table provides an overview of the measurements generated by the respective methodologies.
人体测量指标和指数以及基于仪器的工具用于评估脂肪组织含量和分布的总结。左侧面板总结了最常用的人体测量指标，如 BMI、腰围和颈围测量以及矢状腹径（SAD）。它还提供了用于计算肥胖指数的公式。右侧面板概述了用于评估脂肪组织含量和分布的基于仪器的工具。这些工具包括 BIA、空气置换和水置换方法，以及成像技术如 DXA、CT 和 MRI。表格提供了各方法生成的测量结果的概览。

Full size image

Anthropometric indices  人体测量指标

Anthropometric measurements are quantitative measurements of the body used to assess various physical characteristics. These measurements include parameters such as height, weight, circumference of different body parts (such as waist, hip and limbs), skinfold thickness and BMI, which are routinely used to quantify the level of obesity and body composition.
人体测量学测量是用于评估各种身体特征的定量测量。这些测量包括身高、体重、不同身体部位（如腰围、臀围和四肢）的围度、皮褶厚度和 BMI 等参数，这些参数通常用于量化肥胖水平和身体成分。

BMI is a numerical value calculated from an individual’s weight and height. The major benefit of BMI is that it is easy to calculate. Thus, it has been used extensively in clinical practice and obesity research, and is the basis for WHO’s definition of overweight (BMI 25.0–29.9 kg/m²) and obesity (BMI ≥30 kg/m²)⁵². In epidemiological studies, BMI is helpful when investigating the prevalence of obesity in large populations, tracking trends over time and assessing the effect of obesity on health. Clinically, BMI is used to categorize patients; for example, an individual must typically fall within obesity class III (BMI ≥40 kg/m²) to qualify for gastric bypass surgery. Although BMI has been a cornerstone of obesity research and clinical practice for decades, relying solely on BMI has several limitations and can lead to an incomplete or even misleading understanding of the complexities of obesity. For instance, BMI does not consider factors such as muscle mass, meaning that individuals with a high muscle mass might be classified as having overweight or obesity, even though they have low body fat content. For a given BMI, an individual’s body adipose tissue percentage also changes with age, and the rate of this change differs depending on sex, ethnicity and individual differences^53,54,55,56. Most importantly, although BMI is associated with lipid accumulation and metabolic health in large populations, it is insensitive to the actual distribution of body lipids and adipose tissue, which is critically important for assessing an individual’s metabolic health. As a result, BMI has a relatively poor association with cardiometabolic outcomes, and the community is discussing if the BMI measurement has ‘missed the target’ and should be revised^57,58. At the very least, we argue that BMI should only be used in conjunction with other measurements to provide a more comprehensive understanding of the relationship between body weight and health.
BMI 是根据个体的体重和身高计算得出的数值。BMI 的主要优点在于其易于计算。因此，它被广泛应用于临床实践和肥胖研究中，并且是世界卫生组织（WHO）定义超重（BMI 25.0–29.9 kg/m²）和肥胖（BMI ≥30 kg/m²）的基础。在流行病学研究中，BMI 有助于调查大规模人群中肥胖的患病率、追踪时间趋势以及评估肥胖对健康的影响。临床上，BMI 用于对患者进行分类；例如，个体通常必须达到肥胖 III 级（BMI ≥40 kg/m²）才有资格接受胃旁路手术。尽管几十年来 BMI 一直是肥胖研究和临床实践的基石，但仅依赖 BMI 存在一些局限性，可能导致对肥胖复杂性的理解不完整甚至误导。例如，BMI 未考虑肌肉质量等因素，这意味着肌肉质量高的个体可能被归类为超重或肥胖，即使他们的体脂含量较低。 对于给定的 BMI 值，个体的体脂百分比也会随着年龄而变化，且这一变化速率因性别、种族及个体差异而异 ^53,54,55,56 。尤为关键的是，尽管 BMI 在大规模人群中与脂质积累及代谢健康相关联，但它对体内脂质和脂肪组织的实际分布不敏感，而这对评估个体的代谢健康至关重要。因此，BMI 与心脏代谢结果的相关性较弱，学术界正在讨论 BMI 测量是否“偏离了目标”，并应考虑修订 ^57,58 。至少，我们认为 BMI 应与其他测量指标结合使用，以更全面地理解体重与健康之间的关系。

Waist circumference, and waist-to-hip and waist-to-height ratios are superior to BMI in predicting cardiometabolic risk⁵⁹. However, the clinical implementation of these techniques is sometimes challenging. Tension on the measuring tape can be difficult to minimize, and the tape tends to drop below the horizontal plane. Waist circumference is typically measured at the navel, midway between the hip bone and the bottom of the ribs, but after substantial weight loss the belly typically droops, which can displace the navel by several centimetres, making repeated measurements over time inaccurate. Alternatively, waist circumference can be measured at the elbows, which is a more uniform starting point. The hip is often identified at the widest point of the buttock, or by palpating the hip bone, but the latter can be challenging in individuals with obesity. Thus, clinical implementation of waist-to-hip-based measurements can be complex, and inaccurate measurements can affect our interpretation of research trials^60,61. Interestingly, waist circumference is a more reliable predictor of visceral adiposity in men than in women⁵⁹, and thus thigh circumference measurements are sometimes recommended in women⁶².
腰围、腰臀比和腰高比在预测心脏代谢风险方面优于 BMI ⁵⁹ 。然而，这些技术的临床实施有时具有挑战性。测量带上的张力难以最小化，且测量带容易滑落至水平面以下。腰围通常在肚脐处测量，即髋骨与肋骨底部之间的中点，但在显著减重后，腹部通常会下垂，这可能导致肚脐位置偏移数厘米，使得随时间推移的重复测量不准确。或者，可以在肘部测量腰围，这是一个更为一致的起点。臀部通常被识别为臀部最宽处，或通过触诊髋骨来确定，但对于肥胖个体来说，后者可能具有挑战性。因此，基于腰臀比的测量在临床实施中可能较为复杂，不准确的测量可能影响我们对研究试验的解释 ^60,61 。 有趣的是，腰围在预测男性内脏脂肪方面比女性更为可靠 ⁵⁹ ，因此有时建议对女性进行大腿围测量 ⁶² 。

Another comparatively easy anthropomorphic measure is neck circumference, a marker of upper body SAT that can easily differentiate between normal and abnormal adipose distributions. Neck circumference is an ideal simple screening measure, and can identify individuals with overweight or obesity⁶³, and shows a good association with metabolic risks⁶⁴. However, sex-specific differences have been reported, whereby neck circumference was more strongly associated with cardiometabolic risk factors in women than in men, including total and LDL cholesterol, systolic blood pressure or fasting plasma levels of glucose⁶⁵. Neck circumference can be used in children⁶⁶, but shows a better association with BMI or waist circumference in boys and older children than in girls and younger children⁶⁷.
另一个相对简单的人体测量指标是颈围，它是上半身皮下脂肪组织（SAT）的标志，能够轻易区分正常与异常的脂肪分布。颈围是一种理想的简易筛查手段，可用于识别超重或肥胖个体 ⁶³ ，并且与代谢风险显示出良好的关联性 ⁶⁴ 。然而，已有报告指出性别特异性差异，颈围与女性心脏代谢风险因素的关联性较男性更强，这些因素包括总胆固醇、低密度脂蛋白胆固醇、收缩压或空腹血糖水平 ⁶⁵ 。颈围同样适用于儿童 ⁶⁶ ，但在男孩和较大年龄儿童中，其与体重指数（BMI）或腰围的关联性优于女孩和较小年龄儿童 ⁶⁷ 。

A promising alternative to the above-mentioned techniques is the sagittal abdominal diameter (SAD), which measures the anteroposterior diameter of the abdomen. Briefly, the patient should lie on a firm surface and bend the knees to 90° to reduce back arching, after which one can use a sliding-beam caliper to measure the distance from the lower back to the highest point of the abdomen at the umbilicus level during an exhale. Compared with other anthropometric measures, SAD showed the strongest association with visceral adiposity, irrespective of age, sex and the degree of obesity in a 2023 study⁶⁸, and also in a study in the Asian population⁶⁹. The reason SAD provides such an accurate estimate of VAT accumulation is that abdominal SAT falls towards the sides of the belly when the patient is in a horizontal position, while the VAT is constrained by muscle and bone to remain near the midline. The SAD-to-height ratio (SADHtR) is also an excellent anthropometric index, is more strongly associated with intermediate predictors of cardiometabolic disease than BMI or waist-to-hip-based ratios⁷⁰. Currently, there is a lack of prospective studies showing how SAD or SADHtR might predict major disease outcomes or all-cause mortality. However, SAD was included in the NHANES study in 2011–2016 (ref. ⁶⁸), which should make it possible to compare SAD with other adiposity indicators for predicting all-cause mortality during at least 5 years of follow-up. Collectively, we believe that SAD and/or SADHtR could be the best anthropometric indices of visceral adiposity currently available.
上述技术的一个有前景的替代方法是矢状腹径（SAD），它测量腹部的前后径。简而言之，患者应躺在坚硬的表面上，并将膝盖弯曲至 90°以减少背部拱起，然后可以使用滑动梁卡尺在呼气时测量从下背部到脐部水平腹部最高点的距离。与其他人体测量指标相比，SAD 在 2023 年的一项研究 ⁶⁸ 中显示出与内脏脂肪的最强关联，无论年龄、性别和肥胖程度如何，在亚洲人群的研究 ⁶⁹ 中也得到了类似结果。SAD 之所以能如此准确地估计内脏脂肪组织（VAT）的积累，是因为当患者处于水平位置时，腹部皮下脂肪组织（SAT）会向腹部两侧下垂，而 VAT 则被肌肉和骨骼限制在靠近中线处。SAD 与身高之比（SADHtR）也是一个优秀的人体测量指标，与 BMI 或腰臀比相比，它与心脏代谢疾病的中间预测因子有更强的关联 ⁷⁰ 。 目前，尚缺乏前瞻性研究展示 SAD 或 SADHtR 如何预测主要疾病结局或全因死亡率。然而，SAD 已被纳入 2011-2016 年的 NHANES 研究（参考文献 ⁶⁸ ），这将使得在至少 5 年的随访期内，比较 SAD 与其他肥胖指标预测全因死亡率成为可能。总体而言，我们认为 SAD 和/或 SADHtR 可能是目前可用的最佳内脏肥胖人体测量指标。

To measure SAT thickness by ultrasound or skinfold caliper is at first glance relatively easy and non-invasive, but the techniques are difficult to reproduce⁷¹. For instance, different medical staff often apply different pressure on the ultrasound probe, causing variations in the depressing of the adipose tissue. Similarly, skinfold-grasping techniques can produce substantial variations in measurements if applied inconsistently, and thus adherence to identifying, marking and measuring at a defined site is essential, with most care needed in the biceps and triceps skinfold sites⁷². Importantly, it is also challenging to identify the subcutaneous muscle layer in individuals with morbid obesity, and furthermore, skinfold measurements have poor association with body adiposity in children⁷³. Thus, we conclude that these measurements have limited usefulness in the clinic.
通过超声波或皮褶卡尺测量皮下脂肪厚度乍一看相对简单且无创，但这些技术难以重复 ⁷¹ 。例如，不同的医务人员往往对超声探头施加不同的压力，导致脂肪组织受压程度不一。同样，皮褶抓取技术若应用不一致，也会导致测量结果出现显著差异，因此必须在特定部位进行识别、标记和测量，尤其是在肱二头肌和肱三头肌的皮褶部位需要格外注意 ⁷² 。重要的是，在病态肥胖个体中识别皮下肌肉层也具有挑战性，此外，皮褶测量与儿童体脂含量的关联性较差 ⁷³ 。因此，我们得出结论，这些测量方法在临床上的应用价值有限。

With the advent of smartphones, 3D laser-based photonic scanners and machine learning and/or artificial intelligence, a new way of estimating anthropomorphic measures, adipose tissue and muscle mass has evolved that relies on advanced image analysis to assess body composition and adiposity. These methods are designed to replace or improve classic anthropomorphic methods and indices, such as BMI or waist-to-height ratio. Digital image analysis of at least two pictures taken, for example, by a smartphone, can allow good estimates of body adiposity and BMI^74,75. Photonic scanners are a more instrument-heavy method, based on similar principles of digital image analysis, that reconstructs a 3D model of the body surface topology. These instruments use structured light, shine an array of laser light, or use a passive stereo camera approach to automatically calculate anthropomorphic measures, such as BMI, body and VAT mass and skeletal muscle mass^{76,77,78,79,80}. While image analysis from 2D images and 3D photonic scanners is rapid and easy, a limitation of both methods is that they require patients to wear tight-fitting clothes for accurate measurements. Many of these methods also need further validation of their accuracy compared to more established techniques, and a larger pool of participants in longitudinal clinical trials to evaluate their usefulness. Comparison of adiposity and body composition estimates from 3D scanners and smartphones with reference measures derived from ‘gold standard’ methods such as DXA scanners show close correlation of values^81,82, which they might even outperform other reference methods, such as BIA⁸². However, several studies indicate that 3D-derived body adipose tissue estimates can be less precise with increased adiposity⁸³, and can be better correlated in women than in men⁸².
随着智能手机、基于 3D 激光的光子扫描仪以及机器学习和/或人工智能的出现，一种新的估算人体形态测量、脂肪组织和肌肉质量的方法应运而生，该方法依赖于先进的图像分析来评估身体成分和肥胖程度。这些方法旨在取代或改进经典的人体形态测量方法和指数，如 BMI 或腰围身高比。例如，通过智能手机拍摄的至少两张照片进行数字图像分析，可以很好地估计身体肥胖程度和 BMI ^74,75 。光子扫描仪是一种更为依赖仪器的方法，基于类似的数字图像分析原理，重建身体表面拓扑的 3D 模型。这些仪器使用结构光、发射激光阵列或采用被动立体相机方法，自动计算人体形态测量值，如 BMI、身体和内脏脂肪质量以及骨骼肌质量 ^{76,77,78,79,80} 。尽管从 2D 图像和 3D 光子扫描仪进行图像分析既快速又简便，但两种方法的局限性在于，它们要求患者穿着紧身衣物以确保测量的准确性。 许多这些方法还需要与更成熟的技术相比进一步验证其准确性，并在纵向临床试验中扩大参与者群体以评估其效用。通过将 3D 扫描仪和智能手机得出的肥胖度和身体成分估计值与“金标准”方法（如 DXA 扫描仪）得出的参考测量值进行比较，显示出数值的密切相关性 ^81,82 ，甚至可能优于其他参考方法，如 BIA ⁸² 。然而，多项研究表明，随着肥胖度的增加，3D 衍生的身体脂肪组织估计可能不够精确 ⁸³ ，并且在女性中比在男性中相关性更好 ⁸² 。

Body composition assessment
体成分评估

Body composition refers to the amount and distribution of lean tissue (primarily skeletal muscle), adipose tissue and bone, and provides valuable information regarding adipose distribution.
体成分指的是瘦组织（主要是骨骼肌）、脂肪组织和骨骼的数量及分布，提供了关于脂肪分布的有价值信息。

BIA is one of the quickest and easiest methods for predicting body adiposity, where impedance is measured via an electrical current⁸⁴. The body adipose tissue percentage is estimated based on the lean mass, which contains most of the body’s water. However, using BIA devices has several limitations^84,85 and BIA devices tend to underestimate the body adipose tissue percentage in adults, but overestimate it in children⁸⁶. When choosing a device, it is also important to consider the placement of electrodes; for example, between the hands, between the feet, or between the hands and between the feet (octapolar), as the latter is more reliable⁸⁷. Importantly, an electrical current follows the path of least resistance, meaning that if a person has large amounts SAT, the current can instead pass through internal tissues⁸⁸. This is a problem in patients who have morbid obesity, and thus BIA is unreliable when, for instance, monitoring surgically-induced weight loss⁸⁹. However, BIA can be a good option when safety and feasibility are prioritized, such as when monitoring weight gain in children, but it is then critical not to use different devices or electrode layouts interchangeably⁹⁰.
BIA 是预测体脂率最快捷、简便的方法之一，其通过电流测量阻抗 ⁸⁴ 。体脂率基于瘦体重估算，瘦体重包含身体大部分水分。然而，使用 BIA 设备存在若干限制 ^84,85 ，且 BIA 设备往往低估成人的体脂率，却高估儿童的体脂率 ⁸⁶ 。选择设备时，还需考虑电极的放置位置；例如，手与手之间、脚与脚之间，或手与脚之间（八极法），后者更为可靠 ⁸⁷ 。重要的是，电流遵循最小阻力路径，这意味着若个体拥有大量皮下脂肪组织（SAT），电流可能转而通过内部组织 ⁸⁸ 。这对于病态肥胖患者而言是个问题，因此，在监测手术诱导的体重减轻等情况下，BIA 并不可靠 ⁸⁹ 。 然而，当优先考虑安全性和可行性时，如监测儿童体重增长，BIA 可以是一个不错的选择，但此时关键是不能互换使用不同的设备或电极布局 ⁹⁰ 。

Another simple-to-use tool in the clinic is the trademarked BOD POD device that uses air displacement plethysmography to estimate the body density, and subsequently the adipose tissue distribution⁹¹. Factors that can increase measurement variation include facial hair, body temperature, moisture on the body or in the hair and tightness of clothing, and it is critical to account for the volume of air in the lungs⁹¹. Compared with the BOD POD and BIA, hydrostatic weighing is the most accurate method when assessing total fat content; in this method body density is estimated from the difference between body weight in air and body weight in water⁹². Importantly, variations in bone mineral content and hydration pose substantial limitations to all these methods. In addition, these techniques only measure total fat content and not distribution, although some methods attempt to estimate the latter through mathematical formulas.
另一种在诊所中易于使用的工具是注册商标为 BOD POD 的设备，它利用空气置换体积描记法来估计身体密度，进而评估脂肪组织分布 ⁹¹ 。可能增加测量变异的因素包括面部毛发、体温、身体或头发上的水分以及衣物的紧密度，同时考虑肺部空气体积至关重要 ⁹¹ 。与 BOD POD 和生物电阻抗分析（BIA）相比，水下称重在评估总脂肪含量时是最准确的方法；该方法通过比较空气中体重与水中体重的差异来估算身体密度 ⁹² 。重要的是，骨矿物质含量和水分含量的变化对这些方法都构成了显著限制。此外，这些技术仅测量总脂肪含量而非其分布，尽管某些方法尝试通过数学公式来估算后者。

The best way to directly assess body adipose tissue distribution is through tomographic imaging techniques. The most-used is DXA, a 2D imaging technique⁹³. DXA is the gold standard for bone mineral density measurements, but can also estimate total and regional body adipose and lean tissue mass. It utilizes low-energy X-rays, considered generally harmless⁹⁴. As an X-ray passes through the body, its energy is reduced by amounts that depend on the thickness and absorbance characteristics of the tissue (the attenuation coefficient). DXA uses X-rays with two different energy levels. The images can thus be separated into two components (for example, bone and soft tissue)⁹³. This method is often a practical choice due to its high availability in the clinic, and it is quicker and more accurate than densitometry-based methods. However, as it provides a two-dimensional projection, regional volumes are not measured but rather estimated using anatomical models⁹⁵. Therefore, DXA can have limitations in assessing VAT and muscle quantity compared with CT and MRI. The major disadvantage of DXA is that it is not capable of measuring ectopic fat⁹⁶.
直接评估身体脂肪组织分布的最佳方法是通过断层成像技术。最常用的是 DXA，一种二维成像技术 ⁹³ 。DXA 是骨密度测量的金标准，但也能估计全身和局部脂肪及瘦体组织质量。它利用低能量 X 射线，通常被认为是无害的 ⁹⁴ 。当 X 射线穿过身体时，其能量根据组织的厚度和吸收特性（衰减系数）而减少。DXA 使用两种不同能量水平的 X 射线，因此图像可以分离成两个组成部分（例如，骨骼和软组织） ⁹³ 。这种方法因其在诊所中的高可用性而常作为实用选择，且比基于密度测量的方法更快、更准确。然而，由于它提供的是二维投影，区域体积不是直接测量，而是使用解剖模型进行估计 ⁹⁵ 。因此，与 CT 和 MRI 相比，DXA 在评估内脏脂肪和肌肉量方面可能存在局限性。DXA 的主要缺点是无法测量异位脂肪 ⁹⁶ 。

CT measures tissue density based on the attenuation of X-rays and offers precise measurement of tissue density and composition, including adipose tissue, muscle and bone. As a CT scan offers three-dimensional, high-resolution images and volume representations, it can accurately measure ectopic fat content, although it is not as accurate as MRI. A disadvantage of CT is that labour-intensive segmentations of different compartments are required; however, advances in machine learning and AI are beginning to overcome this issue^97,98,99. Importantly, CT involves higher levels of ionizing radiation than DXA⁹⁴. This poses an ethical problem in research studies, particularly in healthy control individuals and children, due to the potential health risks associated with the higher X-ray dosages.
CT 通过 X 射线的衰减测量组织密度，并提供组织密度和成分（包括脂肪组织、肌肉和骨骼）的精确测量。由于 CT 扫描提供三维、高分辨率的图像和体积表示，它可以准确测量异位脂肪含量，尽管其准确性不如 MRI。CT 的一个缺点是需要对不同区域进行劳动密集型的分割；然而，机器学习和人工智能的进步正在开始克服这一问题 ^97,98,99 。重要的是，CT 涉及的离子辐射水平高于 DXA ⁹⁴ 。由于较高的 X 射线剂量可能带来的健康风险，这在研究中，特别是在健康对照个体和儿童中，提出了一个伦理问题。

An MRI scanner uses strong magnetic fields to measure the magnetic properties of nuclei from certain chemical elements, such as hydrogen atoms, in water or lipid, allowing detailed, 3D, whole-body images⁹⁵. Different post-processing techniques are then used to quantify the various adipose and other tissue types. Unlike CT and DXA, MRI does not involve the use of ionizing radiation, thereby allowing longer acquisition times. A limitation of traditional MRI is that performing the scan and analysing the images generated is very time-consuming. However, there are options for automated and semi-automated analysis techniques to assess MRI-based adipose tissue and muscle composition⁹⁵. Collectively, these novel techniques enable the detailed detection and segmentation of different adipose tissue compartments and individual muscle groups, reducing the manual work for analysing a whole-body dataset to a few minutes, rather than hours.
MRI 扫描仪利用强磁场测量水或脂质中某些化学元素（如氢原子）的核磁特性，从而生成详细的、三维的全身图像 ⁹⁵ 。随后，采用不同的后处理技术来量化各种脂肪及其他组织类型。与 CT 和 DXA 不同，MRI 不涉及电离辐射的使用，因此允许更长的采集时间。传统 MRI 的一个局限在于执行扫描和分析生成的图像非常耗时。然而，现有自动化和半自动化分析技术可用于评估基于 MRI 的脂肪组织和肌肉组成 ⁹⁵ 。总体而言，这些新技术能够详细检测和分割不同的脂肪组织区域及个别肌肉群，将分析全身数据集的手动工作量从数小时减少至几分钟。

Utilizing adiposity measurements in clinical practice
在临床实践中利用肥胖测量

As discussed throughout this article, adipose tissue distribution after weight loss is more important than absolute kilograms lost. Therefore, one can argue that end points for weight loss trials would ideally involve tomographic imaging techniques. Indeed, two meta-analyses of randomized clinical trials that investigated the effects of, for example, long-acting glucagon-like peptide 1 receptor agonists on changes in adiposity found that most trials relied on MRI or CT to accurately assess alterations in body adipose tissue distribution^100,101. Unfortunately, the use of such techniques is unfeasible in larger clinical trials that involve tens of thousands of participants and/or take place at locations where large expensive instrumentation remains unaffordable or impractical. Hence, a mix of proven, cost-effective methods such as DXA and SAD and emerging techniques such as digital image analysis might be more suitable when designing large clinical trials.
正如本文所讨论的，减重后脂肪组织的分布比绝对减重公斤数更为重要。因此，可以认为减重试验的理想终点应涉及断层成像技术。事实上，两项随机临床试验的荟萃分析研究了长效胰高血糖素样肽 1 受体激动剂等对脂肪变化的影响，发现大多数试验依赖 MRI 或 CT 来准确评估身体脂肪组织分布的变化 ^100,101 。遗憾的是，在涉及数万名参与者的大型临床试验中，或在大型昂贵仪器仍无法负担或不切实际的地点进行试验时，使用此类技术是不可行的。因此，在设计大型临床试验时，结合使用已证明具有成本效益的方法（如 DXA 和 SAD）和新兴技术（如数字图像分析）可能更为合适。

One could also think to separate this discussion into different areas of use (for example, the clinical setting of routine patient care versus research studies and clinical trials). For instance, although easy to criticize for its inaccuracy, BMI is almost universally used as a screening measure and initial diagnostic tool. It is likely to remain in clinical use as it is cheap, easy to acquire in almost any setting and useful as a ‘first assessment’ of individuals. However, additional anthropometrics do provide extremely useful information and may better reflect changes to adipose tissue distributions. Several studies have evaluated the complexities of measuring waist circumference in the clinical setting¹⁰². We would rather argue for the increased implementation of SAD measurement in the clinic. SAD does not require the removal of clothing and, assuming there is a space for the patients to lie down, it is easier to perform in a standardized manner, as compared with the measurement of waist, hip or neck circumferences.
人们也可以考虑将这一讨论划分为不同的应用领域（例如，常规患者护理的临床环境与研究性研究和临床试验相对比）。举例来说，尽管 BMI 因其不准确性而易于受到批评，但它几乎被普遍用作筛查手段和初步诊断工具。由于其成本低廉、几乎在任何环境下都易于获取，并且作为个体的“初步评估”非常有用，BMI 很可能继续在临床中使用。然而，额外的身体测量确实提供了极其有用的信息，可能更好地反映脂肪组织分布的变化。多项研究已经评估了在临床环境中测量腰围的复杂性 ¹⁰² 。我们更倾向于主张在临床中增加 SAD（矢状腹径）测量的实施。SAD 测量无需脱去衣物，并且假设有空间让患者躺下，与测量腰围、臀围或颈围相比，它更容易以标准化的方式进行。

The more precise measures of adiposity and adipose tissue distribution include CT, MRI and DXA, as previously mentioned. These techniques are usually widely available for research purposes, particularly in affluent nations. Although these technologies are rarely used in the clinical setting of routine patient care, we would still emphasize the importance of their use in measuring clinical trial outcomes. Given the limited usefulness of BIA, and the cumbersome use of hydrostatic weighing, we feel that DXA and MRI are suitable for use in research studies. Concrete examples of the appropriateness of respective adiposity measurement in the clinical setting can be found In Tables 1,2.
如前所述，更精确的肥胖度和脂肪组织分布测量方法包括 CT、MRI 和 DXA。这些技术通常广泛用于研究目的，特别是在富裕国家。尽管这些技术在常规患者护理的临床环境中很少使用，但我们仍强调其在测量临床试验结果中的重要性。鉴于 BIA 的有限用途和水下称重的繁琐使用，我们认为 DXA 和 MRI 适合用于研究。在临床环境中，各自肥胖度测量适用性的具体示例可在表 1、表 2 中找到。

The challenges of measuring adiposity in the clinical setting are profound, but the importance of choosing the right adiposity measuring tool is illustrated in the debate over the ‘obesity-survival paradox’. This term describes a counterintuitive phenomenon whereby increased BMI is protective and is associated with reduced mortality, particularly in women with a BMI of 25–27 kg/m² (refs. ^103,104,105). This fact has led to an extensive research campaign to identify the underlying mechanisms. However, a 2023 study showed that adjusting for waist-to-height ratios instead of BMI largely eliminates the protective association underlying the obesity-survival paradox¹⁰⁶. This finding illustrates the importance of understanding the limitations and applicability of available adiposity measurements to avoid misunderstandings or flawed conclusions.
在临床环境中测量肥胖的挑战是深远的，但在关于“肥胖-生存悖论”的辩论中，选择正确的肥胖测量工具的重要性得到了体现。这一术语描述了一种反直觉的现象，即增加的 BMI 具有保护作用，并与降低的死亡率相关，特别是在 BMI 为 25–27 kg/m²的女性中（参考文献 1）。这一事实引发了一项广泛的研究活动，以确定其潜在机制。然而，2023 年的一项研究表明，用腰围与身高比代替 BMI 进行调整，很大程度上消除了肥胖-生存悖论背后的保护性关联（参考文献 2）。这一发现说明了理解现有肥胖测量方法的局限性和适用性的重要性，以避免误解或得出有缺陷的结论。

In consideration of the above discussion, using only BMI is no longer an option when assessing adiposity. While imaging-based techniques such as DXA or MRI are the most accurate in assessing visceral and ectopic adiposity, they are difficult to implement at scale compared with conventional anthropometric measures. We argue that one of the simplest and best anthropometric measurements of visceral adiposity is SAD. Although first described in the 1980s¹⁰⁷, the obesity field has been slow to adopt the SAD measurement. Why is that? SAD is simple, affordable and easier to implement than waist-to-hip-based measurements. the US National Center for Health Statistics published a clear protocol on how SAD measurements should be implemented in the clinic^108,109, and we would argue for its extended use.
鉴于上述讨论，在评估肥胖时仅使用 BMI 已不再可行。虽然基于成像的技术（如 DXA 或 MRI）在评估内脏和异位脂肪方面最为准确，但与传统的体测量方法相比，它们难以大规模实施。我们认为，内脏脂肪最简单且最佳的体测量指标之一是 SAD。尽管 SAD 在 20 世纪 80 年代首次被描述 ¹⁰⁷ ，但肥胖领域在采用 SAD 测量方面进展缓慢。原因何在？SAD 简单、经济，且比基于腰臀比的测量更易于实施。美国国家卫生统计中心已发布了关于如何在临床中实施 SAD 测量的明确协议 ^108,109 ，我们主张其更广泛的应用。

Another critical discussion point is considering sex and ethnicity as key factors in susceptibility to obesity-related pathophysiology. For instance, ethnicity must be considered when interpreting data from large European and US-based databases (for example, the UK Biobank¹¹⁰) and whether they are applicable in people of other ethnicities. European clinical guidelines could also consider that, for example, Asian people and women with newly developed T2DM are more prone to obesity-induced cardiometabolic disease than white men, and should be prioritized for weight-loss interventions. As outlined in Supplementary Box 1, the need for a presurgical low-calorie diet prior to gastric bypass is also questionable, given the lack of evidence that it reduces perioperative risks and/or postoperative weight loss¹¹¹.
另一个关键的讨论点是将性别和种族视为肥胖相关病理生理易感性的关键因素。例如，在解释来自欧洲和美国大型数据库（如 UK Biobank ¹¹⁰ ）的数据时，必须考虑种族因素，以及这些数据是否适用于其他种族的人群。欧洲临床指南也应考虑到，例如，亚洲人和新发 2 型糖尿病的女性比白人男性更容易患上肥胖引起的心脏代谢疾病，因此应优先考虑进行减重干预。如补充框 1 所述，胃旁路手术前是否需要术前低热量饮食也存在疑问，因为缺乏证据表明它能降低围手术期风险和/或术后体重减轻 ¹¹¹ 。

In conclusion, there is a welcome debate about what constitutes the best adiposity measurements in the clinic. We are confident that the community can develop improved strategies that better reflect individuals’ adiposity distribution, given that the past decade has offered us a much deeper understanding of the importance of this factor in the development of cardiometabolic disease.
总之，关于临床中最佳肥胖测量方法的讨论是受欢迎的。我们相信，鉴于过去十年我们对这一因素在心脏代谢疾病发展中的重要性有了更深入的理解，社区能够制定出更好地反映个体脂肪分布的策略。